CN108710653A - One kind, which is painted, originally reads aloud order method, apparatus and system - Google Patents
One kind, which is painted, originally reads aloud order method, apparatus and system Download PDFInfo
- Publication number
- CN108710653A CN108710653A CN201810439394.9A CN201810439394A CN108710653A CN 108710653 A CN108710653 A CN 108710653A CN 201810439394 A CN201810439394 A CN 201810439394A CN 108710653 A CN108710653 A CN 108710653A
- Authority
- CN
- China
- Prior art keywords
- paint
- information
- paints
- target
- description information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B5/00—Electrically-operated educational appliances
- G09B5/06—Electrically-operated educational appliances with both visual and audible presentation of the material to be studied
- G09B5/065—Combinations of audio and video presentations, e.g. videotapes, videodiscs, television systems
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Educational Technology (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
It is painted the embodiment of the invention discloses one kind and originally reading aloud order method, apparatus and system, this method includes：Obtain the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition；According to respectively can program request paint this picture description information and character description information the audio frequency characteristics are searched for generally obtaining at least one target and paint this information.The technical solution of the embodiment of the present invention, which solves, paints complicated for operation when originally reading aloud program request, needs to learn the problem of painting this title by heart.Even if in the case that program request paint this input information it is indefinite if can be simple and quick completion target paint this program request, improve the usage experience of user.
Description
Technical field
It paints the present embodiments relate to field of computer technology more particularly to one kind and originally reads aloud order method, device and be
System.
Background technology
With the development of preschool education, for exciting child to understand, painting for ability to express originally read aloud equipment and gradually got home
Long and children favors.Equipment is originally read aloud in current painting, and common order method is automatic plays and two kinds of hand dibbling additional.Automatically
Broadcasting is played automatically successively according to preset order after being switched on；Hand dibbling additional be user by paint originally read aloud button in equipment or
Touch screen be manually entered wait for program request paint this title or number, equipment searches according to number input by user and corresponding paints this progress
It reads aloud.
But automatic play can not carry out painting this program request according to the demand of user, and play user is needed to be familiar with manually
The application method for originally reading aloud equipment is painted, and complicated for operation, when user is to painting this vagueness in memory, cannot accurately input and paint this number
Or when title, then the program request for painting this can not be carried out, inconvenience is brought to user's program request.
Invention content
The present invention, which provides one kind and paints, originally reads aloud order method, apparatus and system, and operation is multiple when originally reading aloud program request to solve to paint
It is miscellaneous, it needs to learn the problem of painting this title by heart.Even if in the case that program request paint this input information it is indefinite if can be simple and quick
It completes target and paints this program request, improve the usage experience of user.
Order method is originally read aloud in a first aspect, being painted an embodiment of the present invention provides one kind, this method includes：
Obtain the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition；
According to respectively can program request paint this picture description information and character description information fuzzy search is carried out to the audio frequency characteristics
Rope obtains at least one target and paints this information.
Second aspect, the embodiment of the present invention, which additionally provides one kind, paints and originally reads aloud on-demand device, which includes：
Feature acquisition module, for obtaining the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition；
Paint this search module, for according to respectively can program request paint this picture description information and character description information to the sound
Frequency feature, which is searched for generally obtaining at least one target, paints this information.
The third aspect, the embodiment of the present invention, which additionally provides one kind, paints and originally reads aloud VOD system, which includes：Server and
It paints and originally reads aloud equipment；
The server, for obtaining the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition, and according to each
Can program request paint this picture description information and character description information the audio frequency characteristics are searched for generally obtaining it is at least one
Target paints this information；
Described paint originally reads aloud equipment, and at least one target for receiving the server transport paints this information, from
At least one target paints determination in this information and currently paints this information, and currently paints this resource to server request.
The embodiment of the present invention based on paint originally read aloud equipment acquisition playing speech on demand information audio frequency characteristics, according to can program request paint
This picture description information and character description information is searched for generally, determines that at least one target paints this information, solves
It paints complicated for operation when originally reading aloud program request, needs to learn the problem of painting this title by heart.Even if it is indefinite to paint this input information in program request
In the case of, completion target that also can be simple and quick paints this program request, improves the usage experience of user.
Description of the drawings
In order to clearly illustrate the technical solution of exemplary embodiment of the present, below to required in description embodiment
The attached drawing to be used does a simple introduction.Obviously, the attached drawing introduced is a part of the embodiment of the invention to be described
Attached drawing, rather than whole attached drawings without creative efforts, may be used also for those of ordinary skill in the art
To obtain other attached drawings according to these attached drawings.
Fig. 1 is that a kind of of the offer of the embodiment of the present invention one paints the flow chart for originally reading aloud order method；
Fig. 2 is provided by Embodiment 2 of the present invention a kind of to paint the flow chart for originally reading aloud order method；
Fig. 3 is that a kind of of the offer of the embodiment of the present invention three paints the flow chart for originally reading aloud order method；
Fig. 4 is that a kind of of the offer of the embodiment of the present invention four paints the structure diagram for originally reading aloud on-demand device；
Fig. 5 is that a kind of of the offer of the embodiment of the present invention five paints the structural schematic diagram for originally reading aloud VOD system.
Specific implementation mode
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining the present invention rather than limitation of the invention.It also should be noted that in order to just
Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
Embodiment one
Fig. 1 be the embodiment of the present invention one provide it is a kind of painting the flow chart for originally reading aloud order method, the present embodiment is applicable
The case where equipment carries out painting this program request is originally read aloud by painting in user, this method can paint Ben Lang by provided in an embodiment of the present invention
Read point broadcasting device or system execute, which can be used hardware and/or the mode of software is realized, for example, the device is configurable
It is originally read aloud in equipment in server and/or paint.This method specifically includes：
S101 obtains the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition.
Wherein, it includes painting this title, painting this title description information, paint this to paint and originally read aloud the playing speech on demand information of equipment acquisition
Word content description information and at least one of paint this image content description information.That is user is carrying out painting this point
Sowing time can be inputted with voice and paint this accurate program request of title progress, when user is to wanting painting for program request can when this title is not remembered clearly
To paint the description information (keyword for e.g., painting this title) of this title by memory voice input, paint the description of this word content
Information or at least one of the description information for painting illustration content in this.When user plane to largely can program request paint and originally do not know have
Which this when body selects, this demand information can also be painted by inputting oneself program request and paint this program request, such as：" I wants to listen
This is painted about what is studied English, there are the name of study toy, the also picture of toy in the inside ".
It is analog signal to paint and originally read aloud the playing speech on demand information of equipment acquisition, and time domain waveform only represents acoustic pressure and becomes at any time
The relationship of change, cannot well representative voice feature, therefore, it is necessary to the sound waveform of playing speech on demand information is converted to acoustics
Feature.Specifically, the method for extracting audio frequency characteristics from playing speech on demand information has very much, such as mel-frequency cepstrum coefficient (Mel-
FrequencyCepstralCoefficients, MFCC), linear prediction residue error (LPCC), Multimedia Content Description Interface
(MPEG7) etc., it is preferred that since MFCC is more to meet the acoustical principles of people based on cepstrum, the embodiment of the present invention selects MFCC
Carry out the extraction of audio frequency characteristics.
Optionally, obtain sound paint originally read aloud equipment acquisition playing speech on demand information audio frequency characteristics, if this method be by
It paints and originally reads aloud equipment execution, can paint originally to read aloud after equipment carries out audio feature extraction to be transmitted directly to be used in the equipment
Paint the module of this information search.Can paint originally to read aloud equipment progress audio spy if this method is executed by server
Server is sent to after sign extraction, server paints the audio spy for originally reading aloud equipment transmission by communication module therein to obtain
Sign.
S102, according to respectively can program request paint this picture description information and character description information fuzzy search is carried out to audio frequency characteristics
Rope obtains at least one target and paints this information.
Wherein, can program request paint this picture description information and character description information be advance pair can put read paint this picture and
Word obtain after feature recognition can program request paint this text feature, specifically, picture description information refers to painting in this
All pictures carry out semantic understanding analysis after, to every pictures content generate picture tag and image content description believe
Breath.Character description information refers to painting all words that all words in this on picture or audio resource parse into style of writing
After eigen identification, generation paints this title, paints this title description information and paints the description information of this content.Target paints this
Refer to from it is numerous put read to paint search in this meet playing speech on demand information paint this.Optionally, target paints this information and includes
It paints this title and paints this confidence level, wherein the confidence level for painting this, which is this, paints this matching degree with playing speech on demand information.
Specifically, due to playing speech on demand information input by user be by paint this title, paint this title description information, paint herein
Word content description information and composition of at least one of painting this image content description information, therefore according to playing speech on demand information
Target is carried out when painting this search, can from have it is all can program request paint the data of this picture description information and character description information
Target is carried out in library paints searching for generally for this information.It should be noted that the audio frequency characteristics obtained in S101 cannot be directly used to
Target paints searching for generally for this information, needs that analysis first is identified to audio frequency characteristics, obtains the text feature of audio frequency characteristics, then
Using this article eigen from have it is all can program request paint this picture description information and character description information database in carry out
Target paints searching for generally for this information.
It should be noted that this method both individually originally can read aloud equipment execution by painting, can also individually be held by server
Row can also originally read aloud equipment and execution with server from painting.For example, originally reading aloud operand and the storage of equipment due to painting
Measure it is limited, when can program request paint this it is more when, can divide the work to search work, if playing speech on demand information duration is shorter, say
It is bright be it is input by user paint this title, corresponding search arithmetic amount is smaller, can directly from storage can program request paint this title
In searched, this method is originally read aloud equipment and is quickly searched by painting at this time.If playing speech on demand information duration is longer, illustrate
Input by user should be the description information for painting this title or content, and corresponding search is complex, to operand and deposits
That stores up is more demanding, and equipment is originally read aloud in common painting may cannot be satisfied search need, and this method carries out mould by server at this time
Paste search.
It present embodiments provides one kind and paints and originally read aloud order method, based on painting the playing speech on demand information for originally reading aloud equipment acquisition
Audio frequency characteristics, according to can program request paint this picture description information and character description information searched for generally, determine at least
One target paints this information, solves and paints complicated for operation when originally reading aloud program request, needs to learn the problem of painting this title by heart.Even if point
Broadcast paint this input information it is indefinite in the case of, completion target that also can be simple and quick paints this program request, improves making for user
With experience.
Embodiment two
Fig. 2 it is provided by Embodiment 2 of the present invention it is a kind of painting the flow chart for originally reading aloud order method, this method is in above-mentioned implementation
Example on the basis of further optimize, give can program request paint this picture description information and text description information generating process
And how therefrom to search for the introduction that target paints the concrete condition of this information generally.As shown in Fig. 2, this method includes：
S201, scanning can program request paint this every page content.
It is originally typically to be made of plus a small amount of word a series of pictures to paint and originally read aloud painting for device plays.For system
In each can program request paint the content that originally will scan its every page, the content of every page is typically to be made of an at least width picture,
And it sometimes appear that a small amount of word in picture.
Optionally, while this every page content is painted in scanning, can according to paint this title to the content that scans into
Row classification for example, the same scans content for painting this is divided into one kind, and establishes mapping relations, paints originally fuzzy search in target in this way
Suo Shi, can quickly find that description information is corresponding to paint this title according to mapping relations.
S202 is parsed by the scanning result to every page content, generate can program request paint this picture description information
With text description information.
Scanning can program request paint originally obtaining the result is that a series of picture, further include a small amount of word in some pictures, because
This will further parse scanning result, and specific resolving can utilize the light based on convolutional neural networks
Learn character recognition (Optical Character Recognition, OCR) technology and picture semantic analytic technique analysis scanning knot
The content of every width picture in fruit generates picture OCR text informations, picture tag and picture description information；Then it recycles certainly
Right language processing techniques carry out the same picture OCR text informations for painting this, picture tag and the picture description information of extraction
Filtering and further semantic understanding, generate can program request paint this picture description information and text description information.
Optionally, can by generation can program request paint this picture description information and text description information according to painting this title
Classification be stored in the database for painting this information, searched for generally so that user is rapidly completed when painting this program request.
Include not only S201 to S202 it should be noted that in the perfect more new stage to the search data in database
Further include to painting this content description information and painting the processing of this title delineation information, specifically for painting the processing of this pictorial information
Processing method can with when parsing extraction carried out to the audio resource for painting this obtain painting this description, and to paint this title into
The extraction of row keyword generates the description content for painting this title.Final Ben Wenben description informations of painting are painted by what S202 was generated
This character description information paints this content description information and paints this title delineation information and collectively constitutes.
S203 obtains the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition.
S204 is identified analysis to audio frequency characteristics, obtains the text feature of audio frequency characteristics.
S203 obtain audio frequency characteristics cannot show the corresponding specific content information of the sound well, to
It carries out target and paints this search, it is also necessary to further discriminance analysis is carried out to audio frequency characteristics, specifically to the knowledge of audio frequency characteristics
It Fen Xi not may include following two step：
(1) phonetic feature sent is identified based on the speech recognition algorithm of deep learning, the text identified
Word result；
(2) text is generated using term vector (word2vec) technology, keyword extraction techniques etc. to the text results recognized
Eigen recycles the natural language processing technique based on Recognition with Recurrent Neural Network to carry out semantics recognition to text feature, obtains sound
The text feature of frequency feature.
A kind of method of analysis is identified it should be noted that above-mentioned two step is the embodiment of the present invention, may be used also
With use in the prior art any one audio frequency characteristics identification is parsed into the technical method of text information, herein without limit
It is fixed.
It should be noted that the text feature of the audio frequency characteristics obtained in S204 be with generated in S202 can program request paint this
Picture description information and text description information be corresponding, can be retouched by text feature and picture description information and/or text
It states information progress similarity-rough set and paints this to search for target.
S205, according to text feature in database respectively can program request paint this picture description information and character description information into
Row big data comparison is handled；
Due to stored in database can program request paint originally and have very much, every is painted this and has a large amount of picture description information and text
Word description information, directly from database it is numerous can program request paint and scan in this, workload is larger, takes longer.Therefore
Can according to obtained in S204 text feature with big data compare treatment technology from database magnanimity paint it is complete in this information
Cheng Huiben coarse sizing processes.For example, can program request paint this classification story type, learning type, nursery rhymes type etc., if text feature pair
What is answered is nursery rhymes type, then comparing treatment technology scalping by big data selects painting originally for all nursery rhymes types, is screening
Nursery rhymes type paint and further searched for generally again in this, improve the efficiency of search.
Optionally, according to text feature in database respectively can program request paint this picture description information and character description information
Carry out big data comparison processing, can be calculate stored in obtained text feature and database in S204 can program request paint this
The COS distance of picture description information and character description information obtains the similarity size between them, and COS distance is closer, says
It is bright this can program request paint this with target to paint this similarity bigger.
S206 compares progress target in handling result from big data according to text feature and paints searching for generally for this, and to searching
Rope to target paint at least one target that is calculated of this progress confidence level and paint this information.
Treatment technology is compared in S205 by big data to paint this and carried out coarse sizing the reading of putting in database, this
When need to only be obtained from coarse sizing put to read to paint and carry out target according to text feature in this and paint searching for generally for this.Optionally, if
Input by user is when accurately painting this title, and what is searched at this time is that a target paints this information, if input by user is to paint
When this description information, at this time according to the target that description information searches paint this may just have it is multiple.The process searched for generally
Be exactly text feature in database can program request paint this picture description information and character description information carry out similarity-rough set
Process, using similarity higher than threshold value can program request paint this and painted as target.Therefore at least one target selected paints this
Similarity is different, calculates the target that each is searched out and paints this similarity i.e. confidence level.Target is painted this
Title and its corresponding confidence level paint this information collectively as this.
Optionally, calculate after each target searched paints this confidence level, can according to confidence level to target paint this from
Arrive greatly it is small be ranked up, the target after sequence is painted and originally shows user, optionally, can be by sequence after all targets paint
This all shows user, can also be one threshold value of setting, after confidence level is painted this sequence more than at least one target of threshold value
Show user.User can select according to the ranking results of confidence level best suits painting originally for self-demand.For example, working as user plane
To it is numerous can program request paint originally do not know how selection when, some description informations of oneself demand, system are inputted by voice
Equally can according to the description information from database it is numerous can program request paint to search in this and meet at least the one of user demand
A to paint this, and be ranked up by confidence level, user can know and oneself demand matching degree is highest paints this according to ordering scenario
Which is, and carries out program request.
Present embodiments provide one kind and paint and originally read aloud order method, by advance pair can program request paint this and handle, will be every
It is a can program request paint this picture description information and text description information storage in the database, obtain user input playing speech on demand
After the audio frequency characteristics of information, fuzzy search is carried out according to pre-stored picture description information in database and character description information
Rope determines that at least one target paints this information, though in the case that program request paint this input information it is indefinite if can be simply fast
The completion target of speed paints this program request, improves the usage experience of user.
Embodiment three
Fig. 3 be the embodiment of the present invention three provide it is a kind of painting the flow chart for originally reading aloud order method, the present embodiment is with aforementioned
Based on embodiment, a preferred embodiment is provided, is suitable for selecting different executive agents to carry out according to the duration of sound IP Information On Demand
The case where originally reading aloud program request is painted, as shown in figure 3, this method includes：
S301 is painted and is originally read aloud equipment acquisition playing speech on demand information, and carries out audio feature extraction.
It can be voice acquisition module, such as wheat to paint and originally read aloud the module for acquiring playing speech on demand information input by user in equipment
Gram wind.It after collecting playing speech on demand information input by user, needs to carry out audio feature extraction, can be first illustratively
To treated, voice signal is digitized using MFCC technologies after collected playing speech on demand information progress noise reduction process
Processing, extracts the audio frequency characteristics of playing speech on demand information.
S302, paint originally read aloud equipment judge acquisition playing speech on demand information duration whether be more than time threshold, if so, holding
Otherwise row S303 executes S306.
The duration of playing speech on demand information input by user determine user description paint this relevant information number, for one
Common painting originally is read aloud for equipment, and the configuration of processing unit is not very high, when to paint this relevant information more for user's description
When to carry out searching for generally complexity from database larger, it is possible that the case where arithmetic speed does not catch up with or malfunctions, therefore,
The difference for originally reading aloud equipment needs according to user's IP Information On Demand duration is painted, reasonable arrangement target paints the execution pair of this search work
As.Specifically, when playing speech on demand duration is more than time threshold, executes S303 and target is painted into this search work arranged to service
Device is handled；When playing speech on demand duration is less than or equal to time threshold, executes S306 and directly originally read aloud equipment itself by painting
It scans for.
Optionally, usually it is exactly the voice input of several words when playing speech on demand information is title, the time is usually shorter,
For example, two to three seconds can be completed；And if description information it is typically one section of word or several sections of words input by user, relative time
Will be longer, therefore, can such as it set the shorter of time threshold setting to three seconds.
S303 is painted and is originally read aloud equipment the audio frequency characteristics of extraction are sent to server.
S304, server according to respectively can program request paint this picture description information and character description information to audio frequency characteristics carry out
It searches for obtaining at least one target generally and paints this information and be sent to paint and originally read aloud equipment.
After the audio frequency characteristics for the playing speech on demand information that server receives, unified based process is first carried out, specifically：First
The phonetic feature sent is identified based on the speech recognition algorithm of deep learning, the text results identified；It is right again
The text results recognized generate text feature using term vector (word2vec) technology, keyword extraction techniques etc., then, profit
Semantics recognition is carried out to text feature with the natural language processing technique based on Recognition with Recurrent Neural Network, obtains the text of audio frequency characteristics
Feature, for carrying out subsequent search for generally.
The audio frequency characteristics of the playing speech on demand information received due to server are that duration is more than time threshold, should
It is to paint this description information, and voice description information includes painting this title description information, painting this word content description information
And paint this image content description information.Optionally, can this different description information in three be divided into two classes to handle,
(1) title searches for class generally：Including painting this title description information；(2) content searches for class generally：Including painting the description of this word content
Information and paint this image content description information.
Specifically, the method searched for generally for title can be：Text based on the audio frequency characteristics that based process obtains
Feature from database respectively can program request paint in this character description information to paint this title and paint this title description information (e.g.,
Paint this title, keyword, short word etc.) establish index information in carry out fuzzy search, find the high at least one mesh of matching degree
Mark and draw this.
The method searched for generally for content can be：According to text feature in database respectively can program request paint this picture
Description information and character description information carry out big data comparison processing；According to text feature from big data compare handling result in into
Row target paints searching for generally for this, finds the high at least one target of matching degree and paints this.
Optionally, process is searched for for the same playing speech on demand information generally, search for generally with title
Appearance carries out searching for generally that one of them can be only carried out, and can also both execute.
Due to the target searched for generally paint this usually have it is multiple, in order to find out to allowing user to be better understood by
Target paints this matching relationship between the playing speech on demand information of oneself input, can paint this progress confidence to the target searched
The calculating of degree obtains at least one target and paints this information, and be sent to paint originally read aloud equipment for user to be played it is current
Paint this determination.
S305, paint originally read aloud equipment receive server transport at least one target paint this information, from least one target
It paints determination in this information and currently paints this information.
Paint originally read aloud equipment receive server transmission at least one target paint this information after, can originally be read aloud by painting
Display screen in equipment shows search result to user, can be at least one target that will be searched paint this according to confidence level into
It is shown successively after row sequence, can also be to mark this after each target paints this to paint this corresponding confidence level.User is according to painting Ben Lang
The display result points for reading device display screen select that oneself wants to play to paint, originally read aloud equipment when painting and detect that user's clicks behaviour
After work, the target that user clicks is painted into this conduct and currently paints this, and obtains the related resource identifier for painting this, such as paints this name
The compositions such as title, number, storage address currently paint this information.
S306 is painted and is originally read aloud the identification that equipment to the audio frequency characteristics of extraction currently paint sheet, works as if identifying and successfully determining
Before paint this information.
When the audio frequency characteristics duration of playing speech on demand information is less than or equal to time threshold, just equipment progress is originally read aloud by painting
Search, therefore playing speech on demand information should paint this specific name.Specific paint originally is read aloud equipment and is used according to audio frequency characteristics
Current this identification process of painting of family program request is the operation offline order word recognizer of deep learning, inputs audio frequency characteristics, identifies it
Whether be it is known paint this title, if identify successfully, searched for user's displaying by painting the display screen originally read aloud in equipment and tied
Fruit, and obtain this and paint this related resource identifier, such as paint this title, number, storage address composition and currently paint this information.
Optionally, it does not identify success if painting and originally reading aloud equipment, can be disappeared by reading aloud the display screen set output prompt
Breath reminds user to re-enter.For example, " search failure, please input IP Information On Demand " can be shown on a display screen.In view of painting
Originally the user for reading aloud equipment is children, optionally, message progress voice can be will be prompted to while display reminding message and is broadcast
It puts, improves the usage experience of user.
S307 is painted and is originally read aloud equipment and currently paint this resource to server request according to this information is currently painted and read aloud.
Due to painting the limited storage space originally read aloud and set, usually can program request paint this audio message and be stored in server
In, therefore, determine that after currently painting this information, meeting basis currently paints this information ask this to paint this letter to server when reading aloud to set
Cease corresponding audio resource, server, which receives after request, to be sent to this audio resource of painting found to paint originally to read aloud and set
It is standby, it paints and originally reads aloud the broadcasting that equipment carries out currently painting at this time sheet.
It should be noted that the method that S301, S302, S306 and S307 are constituted is suitable for playing speech on demand input by user
Information is to read aloud after equipment receives playing speech on demand information the case where painting this title and itself carry out painting this search；S301-S305
And it is the case where painting this description information, to be connect by server that the method for S307 compositions, which is suitable for playing speech on demand information input by user,
It carries out painting searching for generally for this after receiving playing speech on demand information.Optionally, playing speech on demand information input by user is to paint this description
It is two kinds that the case where information, which is divided into,：(1) user is to painting this title vagueness in memory, and painting for this part title or incorrect pronunciations is painted in input
This title；(2) user only remembers to paint the approximate contents of this approximate contents, illustration, for example, input is painted perhaps internal one in this
A little key persons, the information such as sentence, or only know oneself rough demand, without specific program request target.Both above-mentioned feelings
Condition
It present embodiments provides one kind and paints and originally read aloud order method, to playing speech on demand information input by user according to duration point
Dispensing server or paint originally reads aloud equipment and paint this search, and no matter whether playing speech on demand information input by user is clear, all
Target can efficiently be completed and paint this program request, improve the usage experience of user.
Example IV
Fig. 4 be that the embodiment of the present invention four provides it is a kind of painting the structure diagram for originally reading aloud on-demand device, the device is executable
What any embodiment of the present invention was provided, which paint, originally reads aloud order method, has the corresponding function module of execution method and beneficial to effect
Fruit.As shown in figure 4, the device includes：
Feature acquisition module 401, for obtaining the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition；
Paint this search module 402, for according to respectively can program request paint this picture description information and character description information to institute
It states audio frequency characteristics and is searched for generally obtaining at least one target and paint this information.
It present embodiments provides one kind and paints and originally read aloud on-demand device, by based on painting the playing speech on demand for originally reading aloud equipment acquisition
The audio frequency characteristics of information, according to can program request paint this picture description information and character description information searched for generally, determine
At least one target paints this information, solves and paints complicated for operation when originally reading aloud program request, needs to learn the problem of painting this title by heart.Even if
Paint that this input information is indefinite in program request, completion target that also can be simple and quick paints this program request, improves user
Usage experience.
Further, above-mentioned apparatus further includes：
Scan module, for scan it is described can program request paint this every page content；
Information generating module, for being parsed by the scanning result to every page content, described in generation can program request paint
This picture description information and text description information.
Further, above-mentioned this search module 402 of painting includes：
Discriminance analysis unit, for analysis to be identified to the audio frequency characteristics, the text for obtaining the audio frequency characteristics is special
Sign；
Data pre-processing unit, for according to the text feature in database respectively can program request paint this picture describe letter
Breath and character description information carry out big data comparison processing；
Searching order unit paints this for comparing progress target in handling result from big data according to the text feature
It searches for generally, and at least one target that is calculated that this progress confidence level is painted to the target searched paints this information.
Optionally, if the duration of playing speech on demand information is more than time threshold, the present embodiment described device is configured at service
In device；Otherwise, which is configured to paint and originally read aloud in equipment.
If the device is configured in server, which further includes communication module, and at least one mesh is obtained for that will search for
It marks and draws this information and is sent to and read aloud equipment.
Paint at this time originally read aloud equipment receive communication module transmission at least one target paint this information, from least one
A target paints determination in this information and currently paints this information, and currently paints this resource to server request.
It should be noted that the device can be only configured in server, carrying out target by server paints this search,
It only can be configured to paint and originally read aloud in equipment, it, can also be same by the device by painting the search originally read aloud equipment progress target and paint sheet
When be configured to paint and originally read aloud in equipment and server, carried out target with equipment and server from painting originally to read aloud and painted this search.
It is worth noting that, above-mentioned paint in the embodiment for originally reading aloud on-demand device, included each unit and module are only
It is divided according to function logic, but is not limited to above-mentioned division, as long as corresponding function can be realized；Example
Such as, which can only include acquisition module and processing module, and acquisition module realizes the acquisition of audio frequency characteristics；Processing modules implement
Can program request paint the generation of this information and paint the correlation functions such as this lookup with target.In addition, the specific name of each functional unit also only
It is the protection domain that is not intended to restrict the invention for the ease of mutually distinguishing.
Embodiment five
Fig. 5 be that the embodiment of the present invention five provides it is a kind of painting the structure diagram for originally reading aloud VOD system, the system is executable
The method that any embodiment of the present invention is provided reaches corresponding advantageous effect, this, which is painted, originally reads aloud VOD system 50 and include：Service
It device 501 and paints and originally reads aloud equipment 502.
Server 501, for obtaining the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud the acquisition of equipment 502, and foundation
Respectively can program request paint this picture description information and character description information audio frequency characteristics are searched for generally to obtain at least one mesh
Mark and draw this information；
It paints and originally reads aloud equipment 502, at least one target for receiving the transmission of server 501 paints this information, from least one
A target paints determination in this information and currently paints this information, and asks currently to paint this resource to server 501.
VOD system is originally read aloud in painting for the present embodiment, by based on the sound for painting the playing speech on demand information for originally reading aloud equipment acquisition
Frequency feature, according to can program request paint this picture description information and character description information searched for generally, determine at least one
Target paints this information, solves and paints complicated for operation when originally reading aloud program request, needs to learn the problem of painting this title by heart.Even if being painted in program request
In the case of this input information is indefinite, completion target that also can be simple and quick paints this program request, and improve user uses body
It tests.
Note that above are only presently preferred embodiments of the present invention and institute's application technology principle.It will be appreciated by those skilled in the art that
The present invention is not limited to specific embodiments described here, can carry out for a person skilled in the art it is various it is apparent variation,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out to the present invention by above example
It is described in further detail, but the present invention is not limited only to above example, without departing from the inventive concept, also
May include other more equivalent embodiments, and the scope of the present invention is determined by scope of the appended claims.
Claims (10)
1. one kind, which is painted, originally reads aloud order method, which is characterized in that including：
Obtain the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition；
According to respectively can program request paint this picture description information and character description information the audio frequency characteristics search for generally
This information is painted at least one target.
2. according to the method described in claim 1, it is characterized in that, according to respectively can program request paint this picture description information and word
Description information is searched for generally obtaining before at least one target paints this information to the audio frequency characteristics, further includes：
Described in scanning can program request paint this every page content；
Parsed by the scanning result to every page content, described in generation can program request paint this picture description information and text
Description information.
3. if according to the method described in claim 1, it is characterized in that, the duration of the playing speech on demand information is more than time threshold
Value, then the executive agent of the method is server；Otherwise, the executive agent of the method is that described paint originally reads aloud equipment.
4. according to the method described in claim 1, it is characterized in that, if the executive agent of the method is server, foundation
Respectively can program request paint this picture description information and character description information searched for obtaining generally at least one to the audio frequency characteristics
After a target paints this information, further include：
It is described paint originally to read aloud equipment and receive at least one target of the server transport paint this information, from described at least one
A target paints determination in this information and currently paints this information, and currently paints this resource to server request.
5. according to the method described in claim 1, it is characterized in that, according to respectively can program request paint this picture description information and word
Description information is searched for generally obtaining at least one target to the audio frequency characteristics paints this information, including：
Analysis is identified to the audio frequency characteristics, obtains the text feature of the audio frequency characteristics；
According to the text feature in database respectively can program request paint this picture description information and character description information carry out it is big
Comparing processing；
It is compared from big data according to the text feature and carries out target in handling result and paint this search for generally, and to searching
At least one target that is calculated that target paints this progress confidence level paints this information.
6. one kind, which is painted, originally reads aloud on-demand device, which is characterized in that including：
Feature acquisition module, for obtaining the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition；
Paint this search module, for according to respectively can program request paint this picture description information and character description information to the audio spy
Sign, which is searched for generally obtaining at least one target, paints this information.
7. device according to claim 6, which is characterized in that described device further includes：
Scan module, for scan it is described can program request paint this every page content；
Information generating module, for being parsed by the scanning result to every page content, described in generation can program request paint this
Picture description information and text description information.
8. device according to claim 6, which is characterized in that if the duration of the playing speech on demand information is more than time threshold
Value, then described device is configured in server；Otherwise, described device is configured to paint and originally read aloud in equipment.
9. device according to claim 6, which is characterized in that described this search module of painting includes：
Discriminance analysis unit obtains the text feature of the audio frequency characteristics for analysis to be identified to the audio frequency characteristics；
Data pre-processing unit, for according to the text feature in database respectively can program request paint this picture description information and
Character description information carries out big data comparison processing；
Searching order unit paints the fuzzy of this for comparing progress target in handling result from big data according to the text feature
Search, and at least one target that is calculated that this progress confidence level is painted to the target searched paints this information.
10. one kind, which is painted, originally reads aloud VOD system, which is characterized in that the system comprises server and paint and originally read aloud equipment；
The server, for obtaining the audio frequency characteristics for painting the playing speech on demand information for originally reading aloud equipment acquisition, and according to respectively can point
It broadcasts the picture description information for painting this and character description information searches for the audio frequency characteristics generally to obtain at least one target
Paint this information；
Described paint originally reads aloud equipment, and at least one target for receiving the server transport paints this information, from described
At least one target paints determination in this information and currently paints this information, and currently paints this resource to server request.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810439394.9A CN108710653B (en) | 2018-05-09 | 2018-05-09 | On-demand method, device and system for reading book |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810439394.9A CN108710653B (en) | 2018-05-09 | 2018-05-09 | On-demand method, device and system for reading book |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108710653A true CN108710653A (en) | 2018-10-26 |
| CN108710653B CN108710653B (en) | 2021-05-07 |
Family
ID=63867886
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810439394.9A Active CN108710653B (en) | 2018-05-09 | 2018-05-09 | On-demand method, device and system for reading book |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108710653B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032661A (en) * | 2019-03-13 | 2019-07-19 | 上海趣威文化发展有限公司 | The point for repeating two dimensional code code value based on speech recognition reads control method, device and server |
| CN110287364A (en) * | 2019-06-28 | 2019-09-27 | 合肥讯飞读写科技有限公司 | Voice search method, system, equipment and computer readable storage medium |
| CN111523343A (en) * | 2019-01-16 | 2020-08-11 | 北京字节跳动网络技术有限公司 | Reading interaction method, device, equipment, server and storage medium |
| CN111597906A (en) * | 2020-04-21 | 2020-08-28 | 云知声智能科技股份有限公司 | Method and system for quickly identifying picture book by combining character information |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101464896A (en) * | 2009-01-23 | 2009-06-24 | 安徽科大讯飞信息科技股份有限公司 | Voice fuzzy retrieval method and apparatus |
| US8015196B2 (en) * | 2007-06-18 | 2011-09-06 | Geographic Services, Inc. | Geographic feature name search system |
| CN102831106A (en) * | 2012-08-27 | 2012-12-19 | 腾讯科技（深圳）有限公司 | Electronic document generation method of mobile terminal and mobile terminal |
| CN103020119A (en) * | 2012-11-16 | 2013-04-03 | 北京北森测评技术有限公司 | Conversion method, device and system for converting paper edition resume into electronic edition resume |
| CN103365988A (en) * | 2013-07-05 | 2013-10-23 | 百度在线网络技术（北京）有限公司 | Method and device for loud reading pictures and characters of mobile terminal and mobile terminal |
| US20150310012A1 (en) * | 2012-12-12 | 2015-10-29 | Odd Concepts Inc. | Object-based image search system and search method thereof |
| CN105488135A (en) * | 2015-11-25 | 2016-04-13 | 广州酷狗计算机科技有限公司 | Live content classification method and device |
| CN105844978A (en) * | 2016-05-18 | 2016-08-10 | 华中师范大学 | Primary school Chinese word learning auxiliary speech robot device and work method thereof |
| CN107506377A (en) * | 2017-07-20 | 2017-12-22 | 南开大学 | This generation system is painted in interaction based on commending system |
| US20180032612A1 (en) * | 2017-09-12 | 2018-02-01 | Secrom LLC | Audio-aided data collection and retrieval |
-
2018
- 2018-05-09 CN CN201810439394.9A patent/CN108710653B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8015196B2 (en) * | 2007-06-18 | 2011-09-06 | Geographic Services, Inc. | Geographic feature name search system |
| CN101464896A (en) * | 2009-01-23 | 2009-06-24 | 安徽科大讯飞信息科技股份有限公司 | Voice fuzzy retrieval method and apparatus |
| CN102831106A (en) * | 2012-08-27 | 2012-12-19 | 腾讯科技（深圳）有限公司 | Electronic document generation method of mobile terminal and mobile terminal |
| CN103020119A (en) * | 2012-11-16 | 2013-04-03 | 北京北森测评技术有限公司 | Conversion method, device and system for converting paper edition resume into electronic edition resume |
| US20150310012A1 (en) * | 2012-12-12 | 2015-10-29 | Odd Concepts Inc. | Object-based image search system and search method thereof |
| CN103365988A (en) * | 2013-07-05 | 2013-10-23 | 百度在线网络技术（北京）有限公司 | Method and device for loud reading pictures and characters of mobile terminal and mobile terminal |
| CN105488135A (en) * | 2015-11-25 | 2016-04-13 | 广州酷狗计算机科技有限公司 | Live content classification method and device |
| CN105844978A (en) * | 2016-05-18 | 2016-08-10 | 华中师范大学 | Primary school Chinese word learning auxiliary speech robot device and work method thereof |
| CN107506377A (en) * | 2017-07-20 | 2017-12-22 | 南开大学 | This generation system is painted in interaction based on commending system |
| US20180032612A1 (en) * | 2017-09-12 | 2018-02-01 | Secrom LLC | Audio-aided data collection and retrieval |
Non-Patent Citations (1)
| Title |
|---|
| 刘小华: """互联网+"时代的少儿绘本营销"", 《传播与版权》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111523343A (en) * | 2019-01-16 | 2020-08-11 | 北京字节跳动网络技术有限公司 | Reading interaction method, device, equipment, server and storage medium |
| CN111523343B (en) * | 2019-01-16 | 2022-04-01 | 北京字节跳动网络技术有限公司 | Reading interaction method, device, equipment, server and storage medium |
| CN110032661A (en) * | 2019-03-13 | 2019-07-19 | 上海趣威文化发展有限公司 | The point for repeating two dimensional code code value based on speech recognition reads control method, device and server |
| CN110287364A (en) * | 2019-06-28 | 2019-09-27 | 合肥讯飞读写科技有限公司 | Voice search method, system, equipment and computer readable storage medium |
| CN111597906A (en) * | 2020-04-21 | 2020-08-28 | 云知声智能科技股份有限公司 | Method and system for quickly identifying picture book by combining character information |
| CN111597906B (en) * | 2020-04-21 | 2023-12-19 | 云知声智能科技股份有限公司 | Quick drawing recognition method and system combined with text information |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108710653B (en) | 2021-05-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108710653A (en) | One kind, which is painted, originally reads aloud order method, apparatus and system | |
| EP1349145B1 (en) | System and method for providing information using spoken dialogue interface | |
| CN110459210A (en) | Answering method, device, equipment and storage medium based on speech analysis | |
| KR20130128716A (en) | Foreign language learning system and method thereof | |
| CN111651497B (en) | User tag mining method and device, storage medium and electronic equipment | |
| Alon et al. | Contextual speech recognition with difficult negative training examples | |
| CN108305618B (en) | Voice acquisition and search method, intelligent pen, search terminal and storage medium | |
| CN108319588B (en) | Text emotion analysis system and method and storage medium | |
| CN112182229A (en) | Text classification model construction method, text classification method and device | |
| CN107301164B (en) | Semantic analysis method and device for mathematical formula | |
| JP5045486B2 (en) | Dialogue device and program | |
| KR101410601B1 (en) | Spoken dialogue system using humor utterance and method thereof | |
| CN111554276A (en) | Speech recognition method, device, equipment and computer readable storage medium | |
| CN109346108B (en) | Operation checking method and system | |
| CN114742032A (en) | Interactive data analysis method, apparatus, device, medium, and program product | |
| JP5258056B2 (en) | Question sentence candidate presentation device | |
| CN110570838B (en) | Voice stream processing method and device | |
| CN116821290A (en) | Multitasking dialogue-oriented large language model training method and interaction method | |
| CN112116181B (en) | Classroom quality model training method, classroom quality evaluation method and classroom quality evaluation device | |
| CN115512692B (en) | Voice recognition method, device, equipment and storage medium | |
| CN116403583A (en) | Voice data processing method and device, nonvolatile storage medium and vehicle | |
| CN109063127A (en) | A kind of searching method, device, server and storage medium | |
| CN112071304B (en) | Semantic analysis method and device | |
| CN1275174C (en) | Chinese language input method possessing speech sound identification auxiliary function and its system | |
| CN113468306A (en) | Voice conversation method, device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: Room 508-598, Xitian Gezhuang Town Government Office Building, No. 8 Xitong Road, Miyun District Economic Development Zone, Beijing 101500 Applicant after: BEIJING ROOBO TECHNOLOGY Co.,Ltd. Address before: Room 508-598, Xitian Gezhuang Town Government Office Building, No. 8 Xitong Road, Miyun District Economic Development Zone, Beijing 101500 Applicant before: BEIJING INTELLIGENT STEWARD Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210824 Address after: 301-112, floor 3, building 2, No. 18, YANGFANGDIAN Road, Haidian District, Beijing 100038 Patentee after: Beijing Rubu Technology Co.,Ltd. Address before: Room 508-598, Xitian Gezhuang Town Government Office Building, No. 8 Xitong Road, Miyun District Economic Development Zone, Beijing 101500 Patentee before: BEIJING ROOBO TECHNOLOGY Co.,Ltd. |